Although several immunological reagents have been developed (ex: tumor infiltrating lymphocytes, antigen specific tumor vaccines etc.) during the last few years which can recognize and kill melanoma cells and have the potential of curing melanoma patients, most patients are resistant to this form of therapy. This resistance is probably dependent upon a complexity of variables affecting the balance between activation of the immune system and growth of tumor cells in vivo and our limited understanding of them. My work is focused on analyzing variables potentially affecting this balance. During the last year we have further addressed several issues related to this topic, specifically: 1. The quality and quantity of the immune response to peptide-based vaccination. Our data suggest that the timing and number of vaccinations may strongly affect the effectiveness of vaccination and the intensitiy of the immune responses. We are now investigating this issue through a randomized immunization protocol in which patients are treated with immunization either on a weekly or every three week schedule. 2. Various reasons for tumor escape from immune recognition were also explored during the last year. The most important finding in this vein was the conclusion that most tumors do not respond to therapy because the immune response is not sufficient rather than because they adjust their phenotype to circumvent the immune response. However, in case of successful immune therapy, escape mechanisms may play a predominant role in causing tumor recurrence. Several technologies were implemented or developed in our laboratory such as preparation and utilization of soluble HLA/epitope tetramers, TaqMan-based real time PCR, intra-cellular FACS technology and more recently cDNA microarray technology. In particular a novel RNA amplification method was developed that allows proportional expansion of different RNA species for functional genomics studies. AThis new method of RNA aplification allows the utilization of material derived from fine needle aspirates of tumors to be studies in high throughput systems such as micro-arrays. By comparing the expression profile of gene expressed in lesions that did or did not respond to therapy sample through FNA before therapy we noted that lesions that eventually will respond to therapy may be predetermined to do so by a conducive, immune sensitive microenvironment. Presently we are investigating further the effect of various immune treatments in the tumor microenvironment.